How does this skill handle code review feedback?

It follows a strict Read-Understand-Verify-Evaluate-Respond-Implement pattern, prioritizing technical correctness and evidence over 'performative agreement' or social comfort.

Does this skill work with Git workflows?

Yes, it specifically includes protocols for fetching Git SHAs and verifying changes between branches before merge requests are finalized.

What are Verification Gates in this skill?

Verification Gates are mandatory checkpoints that require the agent to run commands and provide fresh evidence—such as test outputs or build logs—before it is allowed to claim a task is complete.

When should I trigger a code-reviewer subagent?

A subagent review should be requested after completing individual tasks in subagent-driven workflows, after finishing major features, or before merging code into the main branch.

Why does the skill discourage saying 'You are absolutely right'?

This is to prevent 'performative agreement,' where an agent agrees with feedback for social reasons rather than technical ones. The skill mandates technical verification before implementing any suggested changes.

Technical Code Review & Verification

Name: Technical Code Review & Verification
Author: unclecatvn

byunclecatvn

•

コラボレーションツール

Enforces rigorous code review protocols and evidence-based verification gates to ensure technical accuracy and prevent false completion claims.

This skill establishes a high-standard code review workflow for Claude, prioritizing technical rigor over performative agreement. It provides structured protocols for receiving feedback with a critical eye, systematically requesting reviews using subagents, and enforcing mandatory 'verification gates' that require fresh evidence before any task is marked as complete. Ideal for subagent-driven development and high-stakes pull requests, it ensures that every claim of success is backed by passing tests and verified requirements, preventing the common pitfall of premature success claims.

主な機能

01Standardized protocol for evaluating and implementing code review feedback

0212 GitHub stars

03Mandatory verification gates requiring fresh logs before completion claims

04Automated review requests via dedicated code-reviewer subagents

05Elimination of performative agreement to ensure technical correctness

06Structured decision-tree for handling unclear or conflicting feedback

ユースケース

01Managing feedback loops between human developers and AI subagents

02Validating major feature implementations before merging to the main branch

03Enforcing evidence-based reporting for bug fixes and task completions

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add unclecatvn/agent-skills code-review

For use in Claude.ai and ChatGPT

Download Skill